AMERICAN MIDLAND NATURALIST. 129 
sent the foliage of ancestral types, a point that has been discussed 
and explained in Quercus, for instance. 
Now in regard to the Cornaceae, Aucuba japonica is the only one 
of which the seedling-stage has been described by Lubbock in his 
comprehensive work on seedlings, and none are mentioned by 
Klebs.* In the seedling of Aucuba the primary root is stout and 
fleshy; the hypocotyl is quite long, and the cotyledons oblong-ovate, 
obtuse, three or five-nerved, shortly petiolate. The first pair of 
leaves are ovate, while the second and third pairs are reduced to 
small, subulate scales preceding the ultimate, which are ample, 
lanceolate-oblong, distinctly serrate-dentate on the upper half, and 
alternately penni-nerved. In Cornus florida} the primary root is 
rather slender, the hypocotyl erect, and the cotyledons green and 
foliaceous; the primary leaves resemble those of the mature tree, 
but are, however, held in a vertical position, and are somewhat 
narrower than the typical leaves. A very similar structure is repre- 
sented by the seedling of Nyssa, or to be more exact of N. sylvatica, 
since the seedlings of the other species are not known so far. In 
this species of /Vyssa (Fig. 1.) the primary root ( R ) is long, very 
slender and amply ramified, especially at the base, beneath the 
hypocotyl (H). This organ, the hypocotyl, is erect, but slender 
and bears two green, oblong cotyledons, which are obtuse, and 
approximately five-nerved; (Fig. 2.) the following internodes are 
stretched, and the leaves of the first season vary from elliptic with 
entire margins to obovate with the margins dentate. The foliage 
of the seedling, thus, differs in a marked degree from that of the 
mature tree, which we remember is mostly oval or obovate, and 
always with the margins entire. In the second season we meet 
with a single, erect shoot with remote leaves and with the primary 
root developed as a strong, very long, deep root; we observed at 
this stage the same variation in foliage, as a matter of fact it recurs 
for several years so long as the plant is a mere shrub, but ceases 
when it becomes a tree. In other words the dentate leaves charact- 
eristic of N. aquatica occur, also, in JV. sylvatica before it reaches 
maturity. It would be interesting to know whether this be the 
same case of N. Ogeche, of which only the entire leaf is recorded. 
This peculiar variation in regard to the foliage, with the mar- 
gins entire or dentate, does not depend upon the position of the 
* Untersuchungen aus d. bot. Inst. Tuebingen I: 536. 1881-1885. 
f Merck's Report for Dec. 1909. 
